ElevenLabs MCP Server

Instructions

⚠️ COST WARNING: This tool makes an API call to ElevenLabs which may incur costs. Only use when explicitly requested by the user.

Args:
    agent_id: ID of the agent to add the knowledge base to.
    knowledge_base_name: Name of the knowledge base.
    url: URL of the knowledge base.
    input_file_path: Path to the file to add to the knowledge base.
    text: Text to add to the knowledge base.

Implementation Reference

• elevenlabs_mcp/server.py:583-595 (registration)

  Registration of the 'add_knowledge_base_to_agent' tool using the @mcp.tool decorator, including description and parameters.

  @mcp.tool(
      description="""Add a knowledge base to ElevenLabs workspace. Allowed types are epub, pdf, docx, txt, html.
  
      ⚠️ COST WARNING: This tool makes an API call to ElevenLabs which may incur costs. Only use when explicitly requested by the user.
  
      Args:
          agent_id: ID of the agent to add the knowledge base to.
          knowledge_base_name: Name of the knowledge base.
          url: URL of the knowledge base.
          input_file_path: Path to the file to add to the knowledge base.
          text: Text to add to the knowledge base.
      """
  )

• elevenlabs_mcp/server.py:596-659 (handler)

  The handler function implements the logic to create a knowledge base document via ElevenLabs API (from URL, file, or text input) and appends it to the specified agent's conversation configuration.

  def add_knowledge_base_to_agent(
      agent_id: str,
      knowledge_base_name: str,
      url: str | None = None,
      input_file_path: str | None = None,
      text: str | None = None,
  ) -> TextContent:
      provided_params = [
          param for param in [url, input_file_path, text] if param is not None
      ]
      if len(provided_params) == 0:
          make_error("Must provide either a URL, a file, or text")
      if len(provided_params) > 1:
          make_error("Must provide exactly one of: URL, file, or text")
  
      if url is not None:
          response = client.conversational_ai.knowledge_base.documents.create_from_url(
              name=knowledge_base_name,
              url=url,
          )
      else:
          if text is not None:
              text_bytes = text.encode("utf-8")
              text_io = BytesIO(text_bytes)
              text_io.name = "text.txt"
              text_io.content_type = "text/plain"
              file = text_io
          elif input_file_path is not None:
              path = handle_input_file(
                  file_path=input_file_path, audio_content_check=False
              )
              file = open(path, "rb")
  
          response = client.conversational_ai.knowledge_base.documents.create_from_file(
              name=knowledge_base_name,
              file=file,
          )
  
      agent = client.conversational_ai.agents.get(agent_id=agent_id)
  
      agent_config = agent.conversation_config.agent
      knowledge_base_list = (
          agent_config.get("prompt", {}).get("knowledge_base", []) if agent_config else []
      )
      knowledge_base_list.append(
          KnowledgeBaseLocator(
              type="file" if file else "url",
              name=knowledge_base_name,
              id=response.id,
          )
      )
  
      if agent_config and "prompt" not in agent_config:
          agent_config["prompt"] = {}
      if agent_config:
          agent_config["prompt"]["knowledge_base"] = knowledge_base_list
  
      client.conversational_ai.agents.update(
          agent_id=agent_id, conversation_config=agent.conversation_config
      )
      return TextContent(
          type="text",
          text=f"""Knowledge base created with ID: {response.id} and added to agent {agent_id} successfully.""",
      )